doc2md. A utility that extracts text from images or PDFs using a local or remote OpenAI-compatible LLM API endpoint with vision-capable multimodal models. For PDFs, each page is rendered to an image and processed sequentially; outputs are concatenated into a single Markdown document.